Food for thought: Microsoft is pushing users and vendors to Macs and Linux
 

Hi,

Microsoft is pushing users and vendors to Macs and Linux is a good read, especially from a Gnu/Linux standpoint.

Microsoft is opening the door for alternative OS by their mind set.

What say you?

There are many ways that a Goliath such as Microsoft can fall, and the belief that they are invincible is certainly one of them.

One keeps hoping for a US government that would be able to stand up to the likes of Microsoft, but that does not seem to be in the cards.

Some forces and trends:
--Linux keeps getting better
--More young people come into the user pool
--The "system architecture" keeps shifting---eg web-based applications, the cloud, tablets, etc.
--profit margins keep getting thinner--motivating companies to cut costs
--ditto for schools, government agencies, etc.

Personally, I don't see a cataclysmic event--with a giant mushroom cloud emanating from Redmond. Rather, I think it will a slow and steady decline---similar, e.g., to that of Kodak.

This point is pretty interesting to me, actually I don't feel that linux is getting better. If you are speaking about GNU/Linux and the distros out there I do agree at all, there are many that are doing a great job but now I'm disappointed with Linux kernel, I feel that they are pushing out new versions too frequently and those versions are marked stable even if they aren't tested at all. There are many 3.2.* version that come out with bugs that aren't supposed to be there with a theoretically mature project like linux.
I still think that something has to be changed in the linux kernel.
Just IMHO.

Member Response
 

Hi,

I agree with some of what you stated. Microsoft has been putting their eggs in one basket as far as OS for consumer products.

As long as there are people who need repair or support for Microsoft, I will be forced to have a install. At some point in time I will slowly decrease the client base locally since my location may change in the future. Now is the time for property investments and this body just cannot handle the winter cold here in Illinois for much longer.

I am sure there will be users who lock step with Microsoft but eventually things will get broken. Remember, Win/286, Win1 or even 3.0. Plus don't forget about CE, ME, Vista or some of the server OS. Loads of excuses for those failures. Old timers will remember the issues with DOS and how Billy got there.

Remember the CP/M issues?

I have made a lot of money from Microsoft users and have no regrets nor feel animosity or remorse. :)

OEM's have been ruining the Windows Users experience for years by including all that crapware when you buy a new computer that just slows it down and makes it unstable. I don't blame microsoft for deciding to build their own hardware. Apple have never had this problem and it shows in user satisfaction figures. I have little sympathy for the vendors: they brought this upon themselves.

Having said that, I have no love for Microsoft either, and I can't see people, and especially professionals, being happy with the new Metro UI any time soon.

Member Response
 

Hi,
Kernel testing is dependent on the user base. Most Gnu/Linux use a stable version. Users that expect the minor RC to be for functional kernel use are fooling with fire. You can use the stable(s) and if you experience problems with the kernel provide feedback instead of wild a$$ FUD or speculation. :)

If you wish to participate with kernel testing then do so. Just do not through out FUD since some users may believe these false statements thus parroting the FUD. :)

Member Response
 

Hi,

@GazL Most of my repairs are due to Crapware or crippleware that the user gets from the OEM, web or a friend. Made loads of bench time to clean things up. I would have one client who never learned that something would happen and then he would drop off for repair. Easy money. He did not care about the charges but would complain about the situation. I told him to stop using just anything from friends or the web. Been running for 6 months now and no calls from him. Maybe I should follow up. Don't think so!

I do use stable versions, as I said I feel that too often the version are marked as stable without proper testing.

Actually I try to provide feedback as better as possible every time i get an issue. When I find an issue I try also to write a patch but with kernel I never had success.

I may want to participate to kernel testing but what I'm trying to say is that i should no be forced to, there are some environment where i need a stable kernel and kernel marked as stable often are not that stable.

Just some personal consideration: I find your post harsh and impolite, maybe I just "don't get it" because English is not my first language or maybe I didn't explained myself very well and resulted impolite for first and that was not my intention. I'm happy that Linux is getting attention (as well as other OS) but if that has to happen because concurrent quality fall it would be sad. I still think that the Linux kernel has to improve releasing. If I have to be honest I liked more it as used to be before 2.6

Member Response
 

Hi,

Sorry, if you feel the tone of my reply is harsh. No intent! My tone and language are direct.

Gnu/Linux & Linux are generally open source. Red Hat has been one of the major supporters for the kernel and other sources like firmware & drivers. While some firmware & drivers are independently developed or manufacture development.

As to hardware combinations and various types to detect, I think the kernel does a very good job. As to drivers, that is not always Linux kernels fault. Mostly because the available manufactures' drivers do need to be reverse engineered or create from scratch. Most people do blame the kernel when it is actually the lack of proper firmware & drivers. Blame the people who do not share nor wish to share openly the hardware specifications or drivers. This is not Gnu/Linux or Linux at fault. Not too many see monetary gain from the time spent developing software firmware or drivers.

I really think you are confusing the issue with the Linux Kernel & current or new firmware or drivers for leading edge or fairly new equipment.

Other potential issue is with legacy hardware with a Gnu/Linux that has not included legacy support. There needs to be a point where things will roll off or removed from support. Linux or Gnu/Linux can be modified or rolled back to provide legacy support. Do that with Microsoft, bet you will waste loads of time.

One other issue for feedback, Linux users should realize that upstream developers do not always reply back. If they do then feel honored, unless of course they are ignoring you for other reasons.

Please do not be so sensitive. This is a technical forum and sometimes members can be terse, myself included. Thick up the skin a bit. There was no intent of verbal abuse or intentional harm. Try a different translator.
HTH!

I don't want to belabor the point about Linux getting better----well, maybe just a bit...;)

I had a brief vacation from serious Linux work---I bought a laptop at home, threw on Ubuntu--and just lived with whatever happened. At work, I got tired of dealing with the IT folks (supported Linux, but only RHEL---yuck) and just ran a Windows box.

Then I retired

Then I moved to the east coast

and THEN--I remembered it was fun and rewarding to try and stay up on the latest in Linux. It was a real revelation to see things (in any distro) that I used to struggle with and now it "just works". I'm back to my favorite distro (ARCH) which definitely has its moments, and ---in general---there is still the occassional opportunity to solve a problem.

I don't think that ANYTHING I have had trouble with lately is traceable to a kernel issue.

Let me make an example of a recent bug in a stable linux kernel. Most of you would say that this is harmless and indeed it's pretty harmless (it's just the first example that come up in my mind) but made me think for ~15 minutes WTF is going on here. As you have noticed it isn't related to firmware\drivers.
My point is how a bug like this one went to a stable branch ? I think that this should not happen.
About drivers i think that they have to be written by the manufacturer of the hardware, that would be a huge boost for linux kernel (more than microsoft taking bad business guesses) and actually some manufacturer do that. I still ask myself why some companies don't release drivers\specifications or they release binary driver(nvidia), are they scared of being copied or?

The Microsoft user-base (that they're really interested in ...) is not going anywhere, because they can't. Nor, most likely, do they want to. They have made a very heavy investment in, say, the dot-Net framework, and it is working for them. It's moving the freight.

I would not be surprised to see a Microsoft Linux Distro LOL... honestly I might even use it... They have really good developers on such an obsolete platform to make it work.

There are several reasons. Have a look at the free radeon drivers for example. While AMD does support the developers with money and specifications there is still no working video decoding and more than sub-optimal power management. This is mostly caused by two things: the fear to reveal third party IP (power management and video decoding using the UVC unit) and the fear to loose control about certain parts of their hardware. In the video chip example, a fully released specification that can be used to develop a driver can also be possibly be used to compromise the secure playback path for BluRay-videos on Windows.

If you look at Nvidia, they simply don't care. Marketshare is what counts for them. If they would not have some larger customers that are using linux with Nvidia's professional cards (Quadro/Tesla) there wouldn't even be a binary driver, IMHO. This may change, now that they have lost a major deal with China because of their driver policies.

it is great to understand the logical reasoning behind these politics. ATI Radeon Cards have given me hell on Linux. The best I could get was on CentOS even with the 2.4 kernel through the Catalyst driver repos. Ubuntu killed my suspend/resume with their Catalyst repo even with the newer kernel.